Air India Runway Excursion
No sign of a preliminary report, but the final report is promised by mid-January 2021, which seems more than a tad ambitious: https://www.civilaviation.gov.in/sit...3_Aug_2020.pdf
Join Date: Jun 2013
Location: India
Age: 85
Posts: 23
Likes: 0
Received 0 Likes
on
0 Posts
Any pilot who has landed in Kozhikode knows that the runway will have standing water during monsoon From this forum I understand that the Captn. has done over 20 landings inthis runway. This being so going to alternate airport would have been safer
Join Date: Apr 2018
Location: Sudbury, Suffolk
Posts: 226
Likes: 0
Received 0 Likes
on
0 Posts
3.2.1 PROBABLE CAUSE
The probable cause of the accident was the non adherence to SOP by the PF, wherein, he continued an unstabilized approach and landed beyond the touchdown zone, half way down the runway, in spite of ‘Go Around’ call by PM which warranted a mandatory ‘Go Around’ and the failure of the PM to take over controls and execute a ‘Go Around’.
3.2.2 CONTRIBUTORY FACTORS
The investigation team is of the opinion that the role of systemic failures as a contributory factor cannot be overlooked in this accident. A large number of similar accidents/incidents that have continued to take place, more so in AIXL, reinforce existing systemic failures within the aviation sector. These usually occur due to prevailing safety culture that give rise to errors, mistakes and violation of routine tasks performed by people operating within the system. Hence, the contributory factors enumerated below include both the immediate causes and the deeper or systemic causes.
(i) The actions and decisions of the PIC were steered by a misplaced motivation to land back at Kozhikode to operate next day morning flight AXB 1373. The unavailability of sufficient number of Captains at Kozhikode was the result of faulty AIXL HR policy which does not take into account operational requirement while assigning permanent base to its Captains. There was only 01 Captain against 26 First Officers on the posted strength at Kozhikode.
(ii) The PIC had vast experience of landing at Kozhikode under similar weather conditions. This experience might have led to over confidence leading to complacency and a state of reduced conscious attention that would have seriously affected his actions, decision making as well as CRM. 209
(iii) The PIC was taking multiple un-prescribed anti-diabetic drugs that could have probably caused subtle cognitive deficits due to mild hypoglycaemia which probably contributed to errors in complex decision making as well as susceptibility to perceptual errors.
(iv) The possibility of visual illusions causing errors in distance and depth perception (like black hole approach and up-sloping runway) cannot be ruled out due to degraded visual cues of orientation due to low visibility and suboptimal performance of the PIC’s windshield wiper in rain.
(v) Poor CRM was a major contributory factor in this crash. As a consequence of lack of assertiveness and the steep authority gradient in the cockpit, the First Officer did not take over the controls in spite of being well aware of the grave situation. The lack of effective CRM training of AIXL resulted in poor CRM and steep cockpit gradient.
(vi) AIXL policies of upper level management have led to a lack of supervision in training, operations and safety practices, resulting in deficiencies at various levels causing repeated human error accidents in AIXL
(vii) The AIXL pilot training program lacked effectiveness and did not impart the requisite skills for performance enhancement. One of the drawbacks in training was inadequate maintenance and lack of periodic system upgrades of the simulator. Frequently recurring major snags resulted in negative training. Further, pilots were often not checked for all the mandatory flying exercises during simulator check sessions by the Examiners.
(viii) The non availability of OPT made it very difficult for the pilots to quickly calculate accurate landing data in the adverse weather conditions. The quick and accurate calculations would have helped the pilots to foresee the extremely low margin for error, enabling them to opt for other safer alternative.
(ix) The scrutiny of Tech Logs and Maintenance Record showed evidence of nonstandard practice of reporting of certain snags through verbal briefing rather than in writing. There was no entry of windshield wiper snag in the Tech log of VT-AXH. Though it could not be verified, but a verbal briefing regarding this issue is highly probable.
(x) The DATCO changed the runway in use in a hurry to accommodate the departure of AIC 425 without understanding the repercussions on recovery of AXB 1344 in tail winds on a wet runway in rain. He did not caution AXB 1344 of prevailing strong tail winds and also did not convey the updated QNH settings.
(xi) Accuracy of reported surface winds for runway 10 was affected by installation of wind sensor in contravention to the laid down criteria in CAR. This was aggravated by frequent breakdown due to poor maintenance.
(xii) The Tower Met Officer (TMO) was not available in the ATC tower at the time of the accident. The airfield was under two concurrent weather warnings and it is mandatory for the TMO to be present to update and inform the fast changing weather variations to enhance air safety. During adverse weather conditions the presence of the TMO in the ATC tower was even more critical.
(xiii) The AAI has managed to fulfil ICAO and DGCA certification requirements at Kozhikode aerodrome for certain critical areas like RESA, runway lights and approach lights. Each of these, in isolation fulfils the safety criteria however, when considered in totality, this left the aircrew of AXB 1344 with little or no margin for error. Although not directly contributory to the accident causation, availability of runway centreline lights would have certainly enhanced the spatial orientation of the PIC.
(xiv) The absence of a detailed proactive policy and clear cut guidelines by the Regulator on monitoring of Long Landings at the time of the accident was another contributory factor in such runway overrun accidents. Long Landing has been major factor in various accidents and incidents involving runway excursion since 2010 and has not been addressed in CAR Section 5, Series F, Part II.
(xv) DGCA did not comprehensively revise CAR Section 5, Series F, Part II Issue I, dated 30 Sep 99 (Rev. on 26 Jul 2017) on ‘Monitoring of DFDR/QAR/PMR Data for Accident/Incident Prevention’ to address the recommendations of the COI of 2010 AIXL Managlore Crash regarding the exceedance limits, resulting in the persisting ambiguities in this matter.
(xvi) DFDR data monitoring for prevention of accidents/incidents is done by AIXL. However 100% DFDR monitoring is not being done, in spite of the provisions laid down in the relevant CAR and repeated audit observations by DGCA. DFDR data monitoring is the most effective tool to identify exceedance and provide suitable corrective training in order to prevent runway accidents like the crash of AXB 1344. However, ATR submitted by AIXL on the said findings were accepted by DGCA year after year without ascertaining its implementation or giving due importance to its adverse implications.
The probable cause of the accident was the non adherence to SOP by the PF, wherein, he continued an unstabilized approach and landed beyond the touchdown zone, half way down the runway, in spite of ‘Go Around’ call by PM which warranted a mandatory ‘Go Around’ and the failure of the PM to take over controls and execute a ‘Go Around’.
3.2.2 CONTRIBUTORY FACTORS
The investigation team is of the opinion that the role of systemic failures as a contributory factor cannot be overlooked in this accident. A large number of similar accidents/incidents that have continued to take place, more so in AIXL, reinforce existing systemic failures within the aviation sector. These usually occur due to prevailing safety culture that give rise to errors, mistakes and violation of routine tasks performed by people operating within the system. Hence, the contributory factors enumerated below include both the immediate causes and the deeper or systemic causes.
(i) The actions and decisions of the PIC were steered by a misplaced motivation to land back at Kozhikode to operate next day morning flight AXB 1373. The unavailability of sufficient number of Captains at Kozhikode was the result of faulty AIXL HR policy which does not take into account operational requirement while assigning permanent base to its Captains. There was only 01 Captain against 26 First Officers on the posted strength at Kozhikode.
(ii) The PIC had vast experience of landing at Kozhikode under similar weather conditions. This experience might have led to over confidence leading to complacency and a state of reduced conscious attention that would have seriously affected his actions, decision making as well as CRM. 209
(iii) The PIC was taking multiple un-prescribed anti-diabetic drugs that could have probably caused subtle cognitive deficits due to mild hypoglycaemia which probably contributed to errors in complex decision making as well as susceptibility to perceptual errors.
(iv) The possibility of visual illusions causing errors in distance and depth perception (like black hole approach and up-sloping runway) cannot be ruled out due to degraded visual cues of orientation due to low visibility and suboptimal performance of the PIC’s windshield wiper in rain.
(v) Poor CRM was a major contributory factor in this crash. As a consequence of lack of assertiveness and the steep authority gradient in the cockpit, the First Officer did not take over the controls in spite of being well aware of the grave situation. The lack of effective CRM training of AIXL resulted in poor CRM and steep cockpit gradient.
(vi) AIXL policies of upper level management have led to a lack of supervision in training, operations and safety practices, resulting in deficiencies at various levels causing repeated human error accidents in AIXL
(vii) The AIXL pilot training program lacked effectiveness and did not impart the requisite skills for performance enhancement. One of the drawbacks in training was inadequate maintenance and lack of periodic system upgrades of the simulator. Frequently recurring major snags resulted in negative training. Further, pilots were often not checked for all the mandatory flying exercises during simulator check sessions by the Examiners.
(viii) The non availability of OPT made it very difficult for the pilots to quickly calculate accurate landing data in the adverse weather conditions. The quick and accurate calculations would have helped the pilots to foresee the extremely low margin for error, enabling them to opt for other safer alternative.
(ix) The scrutiny of Tech Logs and Maintenance Record showed evidence of nonstandard practice of reporting of certain snags through verbal briefing rather than in writing. There was no entry of windshield wiper snag in the Tech log of VT-AXH. Though it could not be verified, but a verbal briefing regarding this issue is highly probable.
(x) The DATCO changed the runway in use in a hurry to accommodate the departure of AIC 425 without understanding the repercussions on recovery of AXB 1344 in tail winds on a wet runway in rain. He did not caution AXB 1344 of prevailing strong tail winds and also did not convey the updated QNH settings.
(xi) Accuracy of reported surface winds for runway 10 was affected by installation of wind sensor in contravention to the laid down criteria in CAR. This was aggravated by frequent breakdown due to poor maintenance.
(xii) The Tower Met Officer (TMO) was not available in the ATC tower at the time of the accident. The airfield was under two concurrent weather warnings and it is mandatory for the TMO to be present to update and inform the fast changing weather variations to enhance air safety. During adverse weather conditions the presence of the TMO in the ATC tower was even more critical.
(xiii) The AAI has managed to fulfil ICAO and DGCA certification requirements at Kozhikode aerodrome for certain critical areas like RESA, runway lights and approach lights. Each of these, in isolation fulfils the safety criteria however, when considered in totality, this left the aircrew of AXB 1344 with little or no margin for error. Although not directly contributory to the accident causation, availability of runway centreline lights would have certainly enhanced the spatial orientation of the PIC.
(xiv) The absence of a detailed proactive policy and clear cut guidelines by the Regulator on monitoring of Long Landings at the time of the accident was another contributory factor in such runway overrun accidents. Long Landing has been major factor in various accidents and incidents involving runway excursion since 2010 and has not been addressed in CAR Section 5, Series F, Part II.
(xv) DGCA did not comprehensively revise CAR Section 5, Series F, Part II Issue I, dated 30 Sep 99 (Rev. on 26 Jul 2017) on ‘Monitoring of DFDR/QAR/PMR Data for Accident/Incident Prevention’ to address the recommendations of the COI of 2010 AIXL Managlore Crash regarding the exceedance limits, resulting in the persisting ambiguities in this matter.
(xvi) DFDR data monitoring for prevention of accidents/incidents is done by AIXL. However 100% DFDR monitoring is not being done, in spite of the provisions laid down in the relevant CAR and repeated audit observations by DGCA. DFDR data monitoring is the most effective tool to identify exceedance and provide suitable corrective training in order to prevent runway accidents like the crash of AXB 1344. However, ATR submitted by AIXL on the said findings were accepted by DGCA year after year without ascertaining its implementation or giving due importance to its adverse implications.
Join Date: Oct 2005
Location: Pacific Rim
Posts: 64
Likes: 0
Received 0 Likes
on
0 Posts
Both chrome and edge browsers warn that this website (Indian AAIB) is unsafe due to expiry of its security certificate. How embarrassing it is for an Indian government agency to have an unsafe website?
Join Date: Oct 2005
Location: Pacific Rim
Posts: 64
Likes: 0
Received 0 Likes
on
0 Posts
THE SEQUENCE OF EVENTS
The aircraft departed from Dubai for Kozhikode at 10:00 UTC as flight AXB 1344 carrying 184 passengers and six crew members.
AXB 1344 made two approaches for landing at Kozhikode. The aircraft carried out a missed approach on the first attempt while coming into land on runway 28. The second approach was on runway 10 and the aircraft landed at 14:10:25 UTC. The aircraft touched down approximately at 4,438 ft on 8,858 ft long runway, in light rain with tailwind component of 15 knots and a ground speed of 165 knots. The aircraft could not be stopped on the runway and this ended in runway overrun. The aircraft exited the runway 10 end at a ground speed of 84 knots and then overshot the RESA, breaking the ILS antennae and a fence before plummeting down the tabletop runway. The aircraft fell to a depth of approximately 110 ft below the runway elevation and impacted the perimeter road that runs just below the tabletop runway, at a ground speed of 41 knots and then came to an abrupt halt on the airport perimeter road just short of the perimeter wall.
There was fuel leak from both the wing tanks; however, there was no postcrash fire. The aircraft was destroyed and its fuselage broke into three sections. Both engines were completely separated from the wings.
The rescue operations were carried out by the ARFF crew on duty with
help of Central Industrial Security Force (CISF) personnel stationed at the airport and several civilians who rushed to the crash site when the accident occurred. Upon receipt of the information about the aircraft crash the district administration immediately despatched fire tenders and ambulances to the crash site. Nineteen passengers were fatally injured and Seventy Five passengers suffered serious injuries in the accident while Ninety passengers suffered minor or no injuries. Both Pilots suffered fatal injuries while one cabin crew was seriously injured and three cabin crew received minor injuries. The rescue operation was completed at 16:45 UTC (22:15 IST).
The captain (59, ATPL, 10,848 hours total, 4,612 hours on type) was pilot flying, the first officer (32, CPL, 1,989 hours total, 1,723 hours on type) was pilot monitoring.
In detail the AAIB described the sequence of events on the overrun:
At 14:07:59 UTC, aircraft captured the glideslope. At 14:08 UTC, ATC cleared AXB 1344 to land on runway 10 reporting latest weather update as visibility of 2000 m in light rain, runway surface wet and winds 250°/08 Kt. At 14:08:29 UTC the crew discussed Landing Flap selections and decided to go for Flap 30 selection instead of the initially discussed Flap 40 due to expected turbulence. Thereafter flap 30 was selected and Landing Checklist was completed at PA 1667 ft.
At 14:09:41 UTC, the aircraft was established on ILS with autopilot and auto throttle engaged at a PA of 633 ft with a CAS of 150 kt, the ground speed was approximately 175 kt with a calculated descent rate of 750 fpm. .
At 500 ft AGL the autopilot was disengaged while the autothrottle remained engaged, the pitch attitude was reduced and the descent rate began to increase, momentarily reaching 1500 fpm. The PM cautioned PF twice for high ROD, which was acknowledged by the PF for ‘correction’. The approach soon became unstabilized, wherein the ROD and glideslope deviation increased beyond the stabilized approach criteria. The aircraft deviated 1.7 dots below the glideslope. This was followed by two EGPWS alerts (caution) for glideslope ‘glideslope... glideslope’. The PF increased the pitch attitude and the descent rate began to decrease, reaching 300 fpm before increasing again to 1000 fpm.
The aircraft crossed the runway threshold at RA of 92 ft and was moving left of centre line with an actual tail wind component of slightly more than 14 Kt and a cross wind component of 6 knots. At this point speed began to gradually increase towards 160 kt CAS and sink rate was gradually arrested as additional thrust was being manually added despite autothrottle command to reduce thrust. The engine power was continuously increased and reached 83% N1. This happened when the aircraft had already gone past the runway threshold by approximately 1363 ft and was at a height of 20 ft RA. The aircraft continued to float above the runway and moved towards the centreline as its lateral deviation was corrected. The engine thrust was reduced and at 3000 ft beyond the threshold, CAS began to decrease towards the approach speed at approximately 15 ft RA. During this time, 07 seconds before touchdown, while the aircraft was at 16 ft RA, the PM tried to catch the attention of the PF by giving a non standard call “Just check it”. At this point, the aircraft was at 2500 ft beyond threshold, 500 ft short of end of touchdown zone. At Kozhikode, end of touchdown zone is marked at a distance of 3000 ft from the threshold by the simple touchdown zone lights on either side of the centre line.
05 seconds before touchdown, the sink rate began to increase gradually towards 12 fps (720 fpm) as the nose was lowered and thrust was reduced. 03 seconds before touchdown, PM again tried to catch the attention of the PF by giving a feeble, uncomfortable call “...Captain” when the aircraft had crossed the end of touchdown zone (3600 ft beyond the threshold). During this time the engine thrust levers reached idle power setting.
Approximately 1 second before touchdown, while the aircraft was crossing 10 ft RA, at 4200 ft beyond the threshold the PM gave a call of “Go around”. There was no response from PF to the “Go-Around” call and he continued with landing. During the flare, the aircraft floated, which resulted in a long landing along with an extended flare of 16 seconds. The aircraft touched down at 4438 ft on the 8858 ft long runway with a CAS of approximately 150 kt and a GS of 165 kt. It was raining at the time of touchdown and the runway condition reported by DATCO was wet.
PF immediately resorted to max manual braking overriding the auto brake selection, auto speed brakes were fully deployed 1.2 seconds after touchdown extending the spoilers. PM gave the standard calls for “SPEED BRAKE UP” and “AUTOBRAKE DISARM” there was no response from the PF to the standard calls given by the PM as per SOP. Autothrottle disengaged automatically 03 seconds after touchdown. The thrust reversers were commanded to deploy 03 seconds after touchdown. They were deployed within 02 seconds after initiating the command and remained in that state for a brief period of approximately 02 seconds with both engines power increasing to 59%N1. There was no call from PM of “REVERSERS NORMAL” as per SOP. None of the standard calls given by PM were acknowledged by the PF.
Before the thrust reversers could take any effect, they were stowed back. While the reversers were being stowed, the aircraft brake pressure was momentarily reduced, decreasing the longitudinal deceleration. This action by the PF coincides with a call by him of “shit”.
Thrust reversers were deployed for the second time 15 seconds after touchdown, when the aircraft was at 8200 ft beyond the threshold, max reverse thrust was commanded and the engine began to spool up. Thrust reversers remained deployed for a period of approximately 07 seconds and by 9100 ft beyond the threshold (paved portion of RESA) as the CAS approached 60 kt, they were stowed back with the engine fan speed (N1) still high.
Two distinct calls from the PF and PM of “shit” were recorded within a gap of one second on the CVR, when the thrust reversers were stowed again and the aircraft was about to leave the runway surface (paved portion of RESA) and enter soft ground.
Speed brakes were stowed back shortly after ‘SPEED BRAKE DO NOT ARM’
light came on. At this time, aircraft had travelled 105 ft in to RESA (soft ground). However, the commanded brake pressure was recorded on the DFDR till the recording stopped. At no stage, after touchdown, were the thrust levers moved forward at any time on the landing roll.
The aircraft did not stop on the runway and this resulted in runway overrun. The aircraft exited the designated end of runway 10 (8858 ft) at a ground speed of 84.5 Kt and then overshot the RESA, broke the ILS antennae and a fence before plummeting down the tabletop runway. The aircraft fell to a depth of approximately 110 ft below the runway elevation and impacted the perimeter road that runs just below the tabletop runway, at a ground speed of 42 Kt. Aircraft came to an abrupt halt on the airport perimeter road, close to the perimeter wall. There was no post accident fire. The impact caused the aircraft to separate into three sections and resulted in 21 fatalities including both pilots.
The AAIB analysed that whatever possible landing scenario they were looking into, it was not possible to stop the aircraft within the runway remaining with the aircraft touching down 4438 feet past the runway threshold. However, had thrust reversers been applied at maximum reverse thrust, a runway friction coefficient of 0.15, the speedbrakes been extended within 1 second after touchdown and brakes engaging within 2 seconds after touch down, the aircraft could have stopped past the runway end, about 9383 feet past the runway threshold, but before the downslope.
The analysis also showed that even a balked landing, a go around being initiated after touchdown yet before thrust reversers being selected, would not resulted in the aircraft reaching lift off height before the runway end.
The AAIB analysed that the aircraft performed the first approach to runway 28. During final approach the captain's windscreen wiper failed about 27 seconds after it was selected on. Nonetheless the captain continued the approach to decision height, then a go around was initiated. The AAIB annotated that the reason for the go around could not be worked out from the evidence available. When the aircraft climbed out, ATC changed the active runway from 28 to 10 in order to accomodate the departure of another aircraft. This changed the headwinds on approach for runway 28 into tailwinds for the approach to runway 10 however. The crew failed to prepare the approach to runway 10 and in particular compute the landing performance. According to FDR data the aircraft experienced a 16 knots tailwind at 30 feet AGL on final approach to runway 10. In addition the crew had selected flaps 30 rather than flaps 40 for the approach due to expected turbulence, the approach speed thus was 8 knots higher than it would have been with flaps 40, in addition additional drag during the roll out was lost. The AAIB wrote:
Landing with Tailwinds on a Wet Runway
AXB 1344 accepted to land on runway 10 in haste without taking into account the implications of their decision. The cockpit crew did not have a discussion or briefing with regard to landing with tailwinds on a wet runway. Lack of importance given to correct landing technique, compounded by massive increase of thrust up to 83% N1, after crossing the threshold, resulted in extended flare and a long landing.
Further the AAIB wrote: "The crew did not consider the unserviceability of the windshield wiper that warranted a diversion to an alternate airfield as mandated by the SOP."
The AAIB analysed the short final:
At time 60 seconds before touchdown, descent rate increased to approximately 1000 fpm. At time 40 seconds before touchdown, the Autopilot was disengaged as the airplane descended through 800 ft PA while the Autothrottle remained engaged. At time 35 seconds before touchdown, the average pitch attitude was reduced and the descent rate began to increase, momentarily reaching 1500 fpm. At time 31 seconds before touchdown (9 seconds after Autopilot was disengaged), the PM also gave a call for “Rate of Descent” and PF acknowledged this call with “Check”. Realising that the PF had not reduced the ROD sufficiently, the PM again gave a call “Rate of Descent... Captain”. At this stage, they were 0.5 NM from touch-down and were still below the GS. The PF responded to the PM’s second call with “Yeah, Yeah...Correcting... Correcting...Correcting”. 29 seconds before touchdown, the crew increased the pitch attitude and the descent rate began to decrease, reaching 300 fpm by time 23 seconds before touchdown, before increasing once again towards 1000 fpm. Simultaneously, glideslope alert (caution) from EGPWS “Glideslope...Glideslope” was heard twice at 14:09:58 UTC. This indicates that the aircraft had deviated from the glideslope by more than 1.3 dot. At this point, the DFDR data revealed that the aircraft had deviated by 1.7 dot below the glideslope. The PM attracted the PF’s attention to this, by a call of “Check”. This was immediately followed by a worried and strained PM’s call of “uh..um”. Thereafter, the descent was arrested and the aircraft came slightly above the glideslope as both engine thrust increased to around 60% N1. While the aircraft was at 92 ft RA, it crossed the threshold of runway 10 with a ground speed of 169 kt and thrust on both engines was just above 61% N1.
At time 16 seconds before touchdown, the sink rate (negative vertical speed about the CG) was arrested. Flare initiation was difficult to discern due to variations in column deflection. By time 10 seconds before touchdown, while the aircraft was at 1363 ft from threshold, sink rate had decreased to nearly 02 fps (120 fpm) after thrust was increased manually by the crew up to 83% N1 despite autothrottle commands to reduce thrust. The manual throttle inputs were verified by comparing throttle resolver angle inputs and conflicting throttle rate commands. For the following 05 seconds, sink rate remained between 02 and 03 fps (120/180 fpm). During this period, radio altitude decreased from approximately 20 ft to 12 ft, indicating that the aircraft floated. At 07 seconds before touchdown, while the aircraft was at 16 ft RA, the PM once again tried to catch the attention of the PF by calling “Just check it”. At this point the aircraft was at 2500 ft beyond the threshold and 500 ft short of end of touchdown zone which at Kozhikode runway is marked by the lights on either side of the centre line. At time 05 seconds before touchdown, the sink rate began to increase gradually towards 12 fps (720 fpm) as the nose was lowered and thrust was reduced. This was followed by a feeble, uncomfortable call of “...Captain” by the PM when the aircraft had crossed the end of touchdown zone (3600 ft beyond the threshold). While the aircraft was crossing 10 ft RA, he gave a definite call of “Go around”, to which there was no response from the PF. The aircraft touched down in less than 01 second after this call.
The PF continued on unstabilized approach to land in spite of caution from PM and EGPWS. Starting from 35 seconds before touchdown (approximately 700 ft PA on approach) till touch down on the runway the stabilised approach criteria as mentioned in Para 25.4 of company OM part A Chapter 25 mandated a Go Around. On crossing the end of touchdown zone, which is clearly marked he once again continued with a long landing. Even after a clear Go-around call, where it was imperative on the part of PF to initiate a missed approach, he did not go-around. This was a gross violation of the SOP. On seeing no response from the PF, it was mandatory for the PM to take over controls and initiate missed approach. The PM did not take over the controls and initiate a missed approach when required to do so.
The AAIB analysed that the captain was suffering from diabeters typ II Mellitus, which was acceptable for flying duties with a proper prescription of medication, which had been issued to the captain. However, the captain not only possessed these prescribed drugs, but also carried 4 other types of diabetes medication. It was obvious he was not following his prescription, all types of drugs found in his possession were also found in his blood sample. These drugs had a serious risk of causing hypoglycaemia (blood sugar levels below required levels). The AAIB wrote this hypoglyaemia could result in "unconsciousness/coma, palpitation, anxiety, sweating, excessive fatigue, nausea, tingling lips, blurred vision, slurred speech or tremors. It may result in only cognitive effects like mental confusion, light headedness and sluggish psychomotor responses. All this can result in decrement of flying performance, which deteriorates further with the complexity of the task at hand."
The AAIB analysed that when the first officer called for a go around, the captain still continued the landing and concluded this indicates the captain completely lost situational awareness. In addition, "Get-Home-Itis" may have contributed to the desire to complete the landing, the AAIB wrote: "The PF was rostered for a scheduled flight for the following day. Any diversion of AXB 1344 flight would have placed the PF in FDTL and he would not have been available to operate the next day’s flight. PF was aware that there were no additional Captains at that base, other than himself to operate that flight. Hence, the PF created a misplaced motivation for himself (to be available for next day’s flight) and did not divert after the wiper was found unserviceable during the first approach and pressed on to land during the second approach (as another ‘missed approach’ would have left him with no option but to divert). According to the statement of the cabin crew, the PIC seemed anxious to return to Kozhikode in time and hence his actions and decisions were steered by a ‘misplaced’ motivation to land back at Kozhikode as scheduled."
In addition the AAIB analysed: "The ‘Steep Authority gradient’ in the cockpit of AXB 1344 acted as a barrier to the crew involvement, restricting the flow of feedback from FO especially with regard to threat analysis and problem solving. Only the most assertive and confident FOs would be able to challenge the authority of PIC. "
AXB 1344 made two approaches for landing at Kozhikode. The aircraft carried out a missed approach on the first attempt while coming into land on runway 28. The second approach was on runway 10 and the aircraft landed at 14:10:25 UTC. The aircraft touched down approximately at 4,438 ft on 8,858 ft long runway, in light rain with tailwind component of 15 knots and a ground speed of 165 knots. The aircraft could not be stopped on the runway and this ended in runway overrun. The aircraft exited the runway 10 end at a ground speed of 84 knots and then overshot the RESA, breaking the ILS antennae and a fence before plummeting down the tabletop runway. The aircraft fell to a depth of approximately 110 ft below the runway elevation and impacted the perimeter road that runs just below the tabletop runway, at a ground speed of 41 knots and then came to an abrupt halt on the airport perimeter road just short of the perimeter wall.
There was fuel leak from both the wing tanks; however, there was no postcrash fire. The aircraft was destroyed and its fuselage broke into three sections. Both engines were completely separated from the wings.
The rescue operations were carried out by the ARFF crew on duty with
help of Central Industrial Security Force (CISF) personnel stationed at the airport and several civilians who rushed to the crash site when the accident occurred. Upon receipt of the information about the aircraft crash the district administration immediately despatched fire tenders and ambulances to the crash site. Nineteen passengers were fatally injured and Seventy Five passengers suffered serious injuries in the accident while Ninety passengers suffered minor or no injuries. Both Pilots suffered fatal injuries while one cabin crew was seriously injured and three cabin crew received minor injuries. The rescue operation was completed at 16:45 UTC (22:15 IST).
The captain (59, ATPL, 10,848 hours total, 4,612 hours on type) was pilot flying, the first officer (32, CPL, 1,989 hours total, 1,723 hours on type) was pilot monitoring.
In detail the AAIB described the sequence of events on the overrun:
At 14:07:59 UTC, aircraft captured the glideslope. At 14:08 UTC, ATC cleared AXB 1344 to land on runway 10 reporting latest weather update as visibility of 2000 m in light rain, runway surface wet and winds 250°/08 Kt. At 14:08:29 UTC the crew discussed Landing Flap selections and decided to go for Flap 30 selection instead of the initially discussed Flap 40 due to expected turbulence. Thereafter flap 30 was selected and Landing Checklist was completed at PA 1667 ft.
At 14:09:41 UTC, the aircraft was established on ILS with autopilot and auto throttle engaged at a PA of 633 ft with a CAS of 150 kt, the ground speed was approximately 175 kt with a calculated descent rate of 750 fpm. .
At 500 ft AGL the autopilot was disengaged while the autothrottle remained engaged, the pitch attitude was reduced and the descent rate began to increase, momentarily reaching 1500 fpm. The PM cautioned PF twice for high ROD, which was acknowledged by the PF for ‘correction’. The approach soon became unstabilized, wherein the ROD and glideslope deviation increased beyond the stabilized approach criteria. The aircraft deviated 1.7 dots below the glideslope. This was followed by two EGPWS alerts (caution) for glideslope ‘glideslope... glideslope’. The PF increased the pitch attitude and the descent rate began to decrease, reaching 300 fpm before increasing again to 1000 fpm.
The aircraft crossed the runway threshold at RA of 92 ft and was moving left of centre line with an actual tail wind component of slightly more than 14 Kt and a cross wind component of 6 knots. At this point speed began to gradually increase towards 160 kt CAS and sink rate was gradually arrested as additional thrust was being manually added despite autothrottle command to reduce thrust. The engine power was continuously increased and reached 83% N1. This happened when the aircraft had already gone past the runway threshold by approximately 1363 ft and was at a height of 20 ft RA. The aircraft continued to float above the runway and moved towards the centreline as its lateral deviation was corrected. The engine thrust was reduced and at 3000 ft beyond the threshold, CAS began to decrease towards the approach speed at approximately 15 ft RA. During this time, 07 seconds before touchdown, while the aircraft was at 16 ft RA, the PM tried to catch the attention of the PF by giving a non standard call “Just check it”. At this point, the aircraft was at 2500 ft beyond threshold, 500 ft short of end of touchdown zone. At Kozhikode, end of touchdown zone is marked at a distance of 3000 ft from the threshold by the simple touchdown zone lights on either side of the centre line.
05 seconds before touchdown, the sink rate began to increase gradually towards 12 fps (720 fpm) as the nose was lowered and thrust was reduced. 03 seconds before touchdown, PM again tried to catch the attention of the PF by giving a feeble, uncomfortable call “...Captain” when the aircraft had crossed the end of touchdown zone (3600 ft beyond the threshold). During this time the engine thrust levers reached idle power setting.
Approximately 1 second before touchdown, while the aircraft was crossing 10 ft RA, at 4200 ft beyond the threshold the PM gave a call of “Go around”. There was no response from PF to the “Go-Around” call and he continued with landing. During the flare, the aircraft floated, which resulted in a long landing along with an extended flare of 16 seconds. The aircraft touched down at 4438 ft on the 8858 ft long runway with a CAS of approximately 150 kt and a GS of 165 kt. It was raining at the time of touchdown and the runway condition reported by DATCO was wet.
PF immediately resorted to max manual braking overriding the auto brake selection, auto speed brakes were fully deployed 1.2 seconds after touchdown extending the spoilers. PM gave the standard calls for “SPEED BRAKE UP” and “AUTOBRAKE DISARM” there was no response from the PF to the standard calls given by the PM as per SOP. Autothrottle disengaged automatically 03 seconds after touchdown. The thrust reversers were commanded to deploy 03 seconds after touchdown. They were deployed within 02 seconds after initiating the command and remained in that state for a brief period of approximately 02 seconds with both engines power increasing to 59%N1. There was no call from PM of “REVERSERS NORMAL” as per SOP. None of the standard calls given by PM were acknowledged by the PF.
Before the thrust reversers could take any effect, they were stowed back. While the reversers were being stowed, the aircraft brake pressure was momentarily reduced, decreasing the longitudinal deceleration. This action by the PF coincides with a call by him of “shit”.
Thrust reversers were deployed for the second time 15 seconds after touchdown, when the aircraft was at 8200 ft beyond the threshold, max reverse thrust was commanded and the engine began to spool up. Thrust reversers remained deployed for a period of approximately 07 seconds and by 9100 ft beyond the threshold (paved portion of RESA) as the CAS approached 60 kt, they were stowed back with the engine fan speed (N1) still high.
Two distinct calls from the PF and PM of “shit” were recorded within a gap of one second on the CVR, when the thrust reversers were stowed again and the aircraft was about to leave the runway surface (paved portion of RESA) and enter soft ground.
Speed brakes were stowed back shortly after ‘SPEED BRAKE DO NOT ARM’
light came on. At this time, aircraft had travelled 105 ft in to RESA (soft ground). However, the commanded brake pressure was recorded on the DFDR till the recording stopped. At no stage, after touchdown, were the thrust levers moved forward at any time on the landing roll.
The aircraft did not stop on the runway and this resulted in runway overrun. The aircraft exited the designated end of runway 10 (8858 ft) at a ground speed of 84.5 Kt and then overshot the RESA, broke the ILS antennae and a fence before plummeting down the tabletop runway. The aircraft fell to a depth of approximately 110 ft below the runway elevation and impacted the perimeter road that runs just below the tabletop runway, at a ground speed of 42 Kt. Aircraft came to an abrupt halt on the airport perimeter road, close to the perimeter wall. There was no post accident fire. The impact caused the aircraft to separate into three sections and resulted in 21 fatalities including both pilots.
The AAIB analysed that whatever possible landing scenario they were looking into, it was not possible to stop the aircraft within the runway remaining with the aircraft touching down 4438 feet past the runway threshold. However, had thrust reversers been applied at maximum reverse thrust, a runway friction coefficient of 0.15, the speedbrakes been extended within 1 second after touchdown and brakes engaging within 2 seconds after touch down, the aircraft could have stopped past the runway end, about 9383 feet past the runway threshold, but before the downslope.
The analysis also showed that even a balked landing, a go around being initiated after touchdown yet before thrust reversers being selected, would not resulted in the aircraft reaching lift off height before the runway end.
The AAIB analysed that the aircraft performed the first approach to runway 28. During final approach the captain's windscreen wiper failed about 27 seconds after it was selected on. Nonetheless the captain continued the approach to decision height, then a go around was initiated. The AAIB annotated that the reason for the go around could not be worked out from the evidence available. When the aircraft climbed out, ATC changed the active runway from 28 to 10 in order to accomodate the departure of another aircraft. This changed the headwinds on approach for runway 28 into tailwinds for the approach to runway 10 however. The crew failed to prepare the approach to runway 10 and in particular compute the landing performance. According to FDR data the aircraft experienced a 16 knots tailwind at 30 feet AGL on final approach to runway 10. In addition the crew had selected flaps 30 rather than flaps 40 for the approach due to expected turbulence, the approach speed thus was 8 knots higher than it would have been with flaps 40, in addition additional drag during the roll out was lost. The AAIB wrote:
Landing with Tailwinds on a Wet Runway
AXB 1344 accepted to land on runway 10 in haste without taking into account the implications of their decision. The cockpit crew did not have a discussion or briefing with regard to landing with tailwinds on a wet runway. Lack of importance given to correct landing technique, compounded by massive increase of thrust up to 83% N1, after crossing the threshold, resulted in extended flare and a long landing.
Further the AAIB wrote: "The crew did not consider the unserviceability of the windshield wiper that warranted a diversion to an alternate airfield as mandated by the SOP."
The AAIB analysed the short final:
At time 60 seconds before touchdown, descent rate increased to approximately 1000 fpm. At time 40 seconds before touchdown, the Autopilot was disengaged as the airplane descended through 800 ft PA while the Autothrottle remained engaged. At time 35 seconds before touchdown, the average pitch attitude was reduced and the descent rate began to increase, momentarily reaching 1500 fpm. At time 31 seconds before touchdown (9 seconds after Autopilot was disengaged), the PM also gave a call for “Rate of Descent” and PF acknowledged this call with “Check”. Realising that the PF had not reduced the ROD sufficiently, the PM again gave a call “Rate of Descent... Captain”. At this stage, they were 0.5 NM from touch-down and were still below the GS. The PF responded to the PM’s second call with “Yeah, Yeah...Correcting... Correcting...Correcting”. 29 seconds before touchdown, the crew increased the pitch attitude and the descent rate began to decrease, reaching 300 fpm by time 23 seconds before touchdown, before increasing once again towards 1000 fpm. Simultaneously, glideslope alert (caution) from EGPWS “Glideslope...Glideslope” was heard twice at 14:09:58 UTC. This indicates that the aircraft had deviated from the glideslope by more than 1.3 dot. At this point, the DFDR data revealed that the aircraft had deviated by 1.7 dot below the glideslope. The PM attracted the PF’s attention to this, by a call of “Check”. This was immediately followed by a worried and strained PM’s call of “uh..um”. Thereafter, the descent was arrested and the aircraft came slightly above the glideslope as both engine thrust increased to around 60% N1. While the aircraft was at 92 ft RA, it crossed the threshold of runway 10 with a ground speed of 169 kt and thrust on both engines was just above 61% N1.
At time 16 seconds before touchdown, the sink rate (negative vertical speed about the CG) was arrested. Flare initiation was difficult to discern due to variations in column deflection. By time 10 seconds before touchdown, while the aircraft was at 1363 ft from threshold, sink rate had decreased to nearly 02 fps (120 fpm) after thrust was increased manually by the crew up to 83% N1 despite autothrottle commands to reduce thrust. The manual throttle inputs were verified by comparing throttle resolver angle inputs and conflicting throttle rate commands. For the following 05 seconds, sink rate remained between 02 and 03 fps (120/180 fpm). During this period, radio altitude decreased from approximately 20 ft to 12 ft, indicating that the aircraft floated. At 07 seconds before touchdown, while the aircraft was at 16 ft RA, the PM once again tried to catch the attention of the PF by calling “Just check it”. At this point the aircraft was at 2500 ft beyond the threshold and 500 ft short of end of touchdown zone which at Kozhikode runway is marked by the lights on either side of the centre line. At time 05 seconds before touchdown, the sink rate began to increase gradually towards 12 fps (720 fpm) as the nose was lowered and thrust was reduced. This was followed by a feeble, uncomfortable call of “...Captain” by the PM when the aircraft had crossed the end of touchdown zone (3600 ft beyond the threshold). While the aircraft was crossing 10 ft RA, he gave a definite call of “Go around”, to which there was no response from the PF. The aircraft touched down in less than 01 second after this call.
The PF continued on unstabilized approach to land in spite of caution from PM and EGPWS. Starting from 35 seconds before touchdown (approximately 700 ft PA on approach) till touch down on the runway the stabilised approach criteria as mentioned in Para 25.4 of company OM part A Chapter 25 mandated a Go Around. On crossing the end of touchdown zone, which is clearly marked he once again continued with a long landing. Even after a clear Go-around call, where it was imperative on the part of PF to initiate a missed approach, he did not go-around. This was a gross violation of the SOP. On seeing no response from the PF, it was mandatory for the PM to take over controls and initiate missed approach. The PM did not take over the controls and initiate a missed approach when required to do so.
The AAIB analysed that the captain was suffering from diabeters typ II Mellitus, which was acceptable for flying duties with a proper prescription of medication, which had been issued to the captain. However, the captain not only possessed these prescribed drugs, but also carried 4 other types of diabetes medication. It was obvious he was not following his prescription, all types of drugs found in his possession were also found in his blood sample. These drugs had a serious risk of causing hypoglycaemia (blood sugar levels below required levels). The AAIB wrote this hypoglyaemia could result in "unconsciousness/coma, palpitation, anxiety, sweating, excessive fatigue, nausea, tingling lips, blurred vision, slurred speech or tremors. It may result in only cognitive effects like mental confusion, light headedness and sluggish psychomotor responses. All this can result in decrement of flying performance, which deteriorates further with the complexity of the task at hand."
The AAIB analysed that when the first officer called for a go around, the captain still continued the landing and concluded this indicates the captain completely lost situational awareness. In addition, "Get-Home-Itis" may have contributed to the desire to complete the landing, the AAIB wrote: "The PF was rostered for a scheduled flight for the following day. Any diversion of AXB 1344 flight would have placed the PF in FDTL and he would not have been available to operate the next day’s flight. PF was aware that there were no additional Captains at that base, other than himself to operate that flight. Hence, the PF created a misplaced motivation for himself (to be available for next day’s flight) and did not divert after the wiper was found unserviceable during the first approach and pressed on to land during the second approach (as another ‘missed approach’ would have left him with no option but to divert). According to the statement of the cabin crew, the PIC seemed anxious to return to Kozhikode in time and hence his actions and decisions were steered by a ‘misplaced’ motivation to land back at Kozhikode as scheduled."
In addition the AAIB analysed: "The ‘Steep Authority gradient’ in the cockpit of AXB 1344 acted as a barrier to the crew involvement, restricting the flow of feedback from FO especially with regard to threat analysis and problem solving. Only the most assertive and confident FOs would be able to challenge the authority of PIC. "
Join Date: Mar 2010
Location: Often in Jersey, but mainly in the past.
Age: 79
Posts: 7,698
Received 102 Likes
on
48 Posts
Skywards747
Likewise Mac/Safari. Sums it all up, really.
Likewise Mac/Safari. Sums it all up, really.
There is good news here. The accident report clearly did not pull any punches' which certainly has not been the been the case for other reports from non Western airline accidents
BUT, has anything changed with respect to DGCA oversight ? The regulator was enviously not paying any attention to this airline.
BUT, has anything changed with respect to DGCA oversight ? The regulator was enviously not paying any attention to this airline.
